Sign in Join

Live chat for Free Pascal, Lazarus, Delphi and every Pascal programmer.

Mailing lists / lazarus / 2010-11

Re: gdb driving me nuts => it's starting a 2nd thread for no apparent reason....

3 messages · 2 people ·

Reply by email Subscribe to this list Your subscription

Read-only archive. Use j and k to move between messages.

Martin
Martin wrote 5 lines
it's the signal issue.... I know how to fix, only matter of time

Actually it probably was the same issue i had..
and probably is fixed now.

The signal is because the IDE tried to interrupt the process, so it could get the line info (blue dots) for the open unit. But at that time the app already stopped at a breakpoint (only the IDE hadn't yet noticed).
So the signal is handled later, when you step over.

There still is a bug, that step-over does exec-continue in such a case. But it will no longer matter (since I moved the line-info to be called while already paused)

There always is (and probably will be) a possibility of this happening, if the user interupts the app, just at the same time it hits a breakpoint.

There also is the issue, that certain actions (set/del breakpoint / switching to unit, that requires line info blue-dots) must interrupt the debugger. and if it does, there is no way to continue an exec-next (step over) since the app can be anywher in some nested subroutine call.... (need to investigate gdb async interface, but it's not yet on all platforms)

Martin

Reply by email

Marc Weustink
to Martin
Martin wrote
(need to investigate gdb async interface, but it's not yet on all platforms)

Be carefull, the async interface isn't implemented on all supported platforms nor supported gdb versions.

Marc

Reply by email

Martin
to Marc Weustink
Marc Weustink wrote 7 lines
Martin wrote: > (need to investigate gdb async interface, but it's not yet on all > platforms) Be carefull, the async interface isn't implemented on all supported platforms nor supported gdb versions.

I know. that part is a longterm idea. Definitely not to be touched before 1.0 => don't want to create to big a mess.

Reply by email